Nick T. Thomopoulos
Illinois Institute of Technology
Abstract
The typical way service level is measured in industry is the demand filled over total demand. Unfilled demand becomes a backorder or lost sales. Lost sales demand is often not known or measured by the management. This paper shows how to estimate the lost sales demand and also shows how to measure an effective service level. Table values are provided for easy reference and three examples show how to apply the results of this paper to industry situations.
Nick T. Thomopoulos, Ph.D.
Illinois Institute of Technology
Nick Z. Malham
FIC Inc.
Mark Spieglan, Ph.D.
FIC Inc.
Abstract
A probabilistic model is developed to measure the difference in forecast accuracy when plant forecasts are based on the shipment history instead of the demand history. This often is the case in the plants. This paper assumes the monthly demands follow a Poisson probability distribution. The paper also assumes the shipments are in lot sizes of quantity q and when the demands accumulate to q, a shipment is made. The probability distribution on the size of monthly shipments is generated. The mean and standard deviation of the monthly shipments are compared to the mean and standard deviation of the monthly demands.
Nick T. Thomopoulos, Ph.D.
Illinois Institute of Technology
Abstract
This paper
gives a comparison between two fundamental methods of determining the safety
stock. The methods are denoted as
Availability and Service Level. Tables
that compare the methods are provided and list how much safety stock in month’s
supply is needed for various situations